Munda in Early NW India

[petegray]

Sanskrit fricatives
> ...Visarga /h./ as a "hard spirant"

There are two h sounds, the voiceless visarga, and the sound transcribed
<h>, which is a voiced sound - so it can't be the usual /h/ sound of
English.   Does that make it a fricative?

The three sibilants are not all allophones of /s/.   One is the direct
reflex of PIE k' (as in as'va horse).   Another prbably began life as an
allophone (mostly becasue of the RUKI rule) but in attested Skt it must be
treated as a phoneme - there are enough minimal pairs.

>> How do we explain .., Iranian where *dh etc became d etc (at least
>> intervocalic)?

I understand (but may be wrong) that the written Iranian d is a fricative.
The voiced aspirates also appear as fricatives in Germanic and Latin, which
develop to stops later.

> My main point was that traditional PIE is short on fricatives and long on
> "laryngeals ... Their characterization
> as "laryngeal" ...

"Laryngeal" is just a label, and not a phonetic description.   The only
evidence we have is the evidence we've got, which allows us to interpret the
"laryngeals" as velars or post-velars.  There is nothing to stop them being
palatals, if you can make it fit the evidence.

Peter